data-structure-protocol

A dependency graph maps project entities by recording UIDs, descriptions, imports, and exports in a .dsp/ directory. It captures the reasons for each connection, enabling traceability and structural reasoning for future codebase changes.

Long-term structural memory is maintained by using DSP to store codebase entities, dependencies, and public APIs in a .dsp/ directory. It records the reasons for every connection, allowing LLM agents to reason about and navigate complex project structures.

You bootstrap project structure by initializing DSP and using the dsp-cli workflow to create objects, functions, and relationships. This records modules and their imports in the .dsp directory, enabling module discovery and dependency analysis.

Yes, a .dsp/ directory is required. DSP applies to codebases that maintain a DSP-tracked structure, using this directory to store the dependency graph and analyze modules, imports, and exports for navigation and dependency analysis.

Yes, DSP captures the reasons for each connection to explain why components are related, enabling traceability and reasoning for future changes. This builds a graph-based structural memory of your project within the .dsp directory.

DSP should not be used if your project lacks a .dsp/ directory or does not require a graph-based structural memory. It is specifically designed for codebases analyzed for modules, imports, and exports that need long-term dependency tracking.